A speckle technique is being developed for full-field displacement analysis of large structures illuminated by sunlight. Potential applications include displacement measurement of civil engineering and geological bodies. A particularly challenging and interesting application is the study of rock debris-covered glaciers where the surface offers an excellent natural speckle pattern. The large object distances involved in the study of glaciers (up to 6km) introduce questions regarding atmospheric effects, lens aberrations and resolution, speckle generation and the suitability of sunlight as an illumination source. This paper describes a study of the feasibility of speckle photography in the presence of these factors. It then presents quantitative results illustrating the effect these factors have on the sensitivity and accuracy of the method. In particular we consider; (1) The spatial frequency content of an incoherent speckle image is dependent on the filtering of object spatial frequencies by a transfer function formed from the combined imaging properties of the lens and the atmosphere. Experimental results characterising the effects of atmosphere on imaging quality are presented; (2) A unique speckle pattern is generated by a certain illumination angle so daily speckle photographs of the glacier surface are taken at a similar time. We consider how much tolerance the technique has to variation in sun angle.

A technique is being developed based on laser and whitelight speckle photography using the Sun as the light source. It is applicable for in-plane displacement analysis of large structures and bodies. For the method to be applicable the surface of the object must have a granular texture of such contrast and definition that it can be resolved by a camera lens. Rough concrete, sand, rock and boulders create a suitable speckle pattern in bright sunlight; the only proviso is that the granular `signature' of the surface does not change. The application of the method to measure the full-field movement of debris covered glaciers is illustrated

A statistical distribution model for point information in speckle photography is established using statistical approach. It is shown that the recognition of information is to recognize the space and direction of the Young's fringe from a random metrological field statistically. Based on the model, a 1-D recovnition technique is presented, in which the homomorphic filtering technique is used to elliminate both the speckle and the diffraction hole effects. Experiments show good accuracy and fast recognition speed.

It is shown that stroboscopic holography can be usefully applied to study the vibration of large machines. It has been demonstrated that a great deal of complementary information can be gained from the stroboscopic technique compared with the normal time-average method, in particular damping characteristics and cyclic idiosyncrasies can be evaluated

It is usual, in speckle photography, to set the optical axis of the photographic lens so that it is perpendicular to the surface being investigated. However if the photographic lens is set with its optical axis at an angle to the object surface then it is shown that the accuracy of separating out the in-plane and out-of-plane displacements can be improved. A further benefit is that the whole visual field of the photographic lens can be studied whether the object itself is flat or has a 3-D surface profile. This means that speckle photography can be applied to objects of any shape and of considerable size. Firstly the principle and theory of the oblique axis method is explained. In particular the method of correct focusing is shown; because of this the technique is not possible with simple 35 mm SLR cameras where lens and film planes do not have the necessary movements for differential focusing, none the less the potential of a mono-rail plate camera can be full realized. Secondly the validity of the technique will be seen by presenting results of tests on a simple structure. Finally it will be shown, using automatic fringe analysis, that the method can be applied to large 3-D structures, such as a machine tool.